Conversational AI
Conversational AI is the set of technologies that lets software hold a dialogue in natural language: it understands the user's requests, maintains the context of the conversation and generates relevant answers. Current systems are based on large language models (LLMs), capable of handling free-form phrasing and multiple languages without predefined paths, unlike rule-based chatbots.
How it works applied to messaging: the customer's message reaches the model along with the context — recent history, behavioral instructions and the company knowledge base — and the model generates the reply that the platform sends in the chat. Quality depends on three factors: the model chosen, the instructions that define its tone and boundaries, and the content it is trained on. Without good content, even the best model improvises.
In commercial use cases, conversational AI covers the first level: questions about products, hours, prices, order status, bookings, at any time and in multiple languages. The limits need governing: the possibility of inaccurate answers, contained with curated knowledge bases, instructions on topics to avoid and handover to an agent in sensitive cases. Meta's policies require transparency about the use of automated systems.
SendApp's conversational AI answers customers on WhatsApp and the other connected channels, trained on your company's content and with configurable handover to an agent.
Frequently asked questions
From theory to practice
Everything you read in this glossary can be done with SendApp: try it free, no credit card required.